© 2012 coreservlets.com and Dima May
Customized Java EE Training: http://courses.coreservlets.com/Hadoop, Java, JSF 2, PrimeFaces, Servlets, JSP, Ajax, jQuery, Spring, Hibernate, RESTful Web Services, Android.
Developed and taught by well-known author and developer. At public venues or onsite at your location.
HDFS - Java API
Originals of Slides and Source Code for Examples:http://www.coreservlets.com/hadoop-tutorial/
© 2012 coreservlets.com and Dima May
Customized Java EE Training: http://courses.coreservlets.com/Hadoop, Java, JSF 2, PrimeFaces, Servlets, JSP, Ajax, jQuery, Spring, Hibernate, RESTful Web Services, Android.
Developed and taught by well-known author and developer. At public venues or onsite at your location.
For live Hadoop training, please see courses at http://courses.coreservlets.com/.
Taught by the author of this Hadoop tutorial. Available at public venues, or customized versions can be held
on-site at your organization.• Courses developed and taught by Marty Hall
– JSF 2, PrimeFaces, servlets/JSP, Ajax, jQuery, Android development, Java 6 or 7 programming, custom mix of topics– Ajax courses can concentrate on 1 library (jQuery, Prototype/Scriptaculous, Ext-JS, Dojo, etc.) or survey several
• Courses developed and taught by coreservlets.com experts (edited by Marty)– Hadoop, Spring, Hibernate/JPA, GWT, SOAP-based and RESTful Web Services
Contact [email protected] for details
Agenda
• Java API Introduction• Configuration• Reading Data• Writing Data• Browsing file system
4
File System Java API
• org.apache.hadoop.fs.FileSystem– Abstract class that serves as a generic file system
representation– Note it’s a class and not an Interface
• Implemented in several flavors– Ex. Distributed or Local
5
FileSystem Implementations
• Hadoop ships with multiple concrete implementations:– org.apache.hadoop.fs.LocalFileSystem
• Good old native file system using local disk(s)
– org.apache.hadoop.hdfs.DistributedFileSystem• Hadoop Distributed File System (HDFS)• Will mostly focus on this implementation
– org.apache.hadoop.hdfs.HftpFileSystem• Access HDFS in read-only mode over HTTP
– org.apache.hadoop.fs.ftp.FTPFileSystem• File system on FTP server
6
FileSystem Implementations
• FileSystem concrete implementations– Two options that are backed by Amazon S3 cloud
• org.apache.hadoop.fs.s3.S3FileSystem• http://wiki.apache.org/hadoop/AmazonS3
– org.apache.hadoop.fs.kfs.KosmosFileSystem• Backed by CloudStore• http://code.google.com/p/kosmosfs
7
FileSystem Implementations
• Different use cases for different concrete implementations
• HDFS is the most common choice– org.apache.hadoop.hdfs.DistributedFileSystem
8
SimpleLocalLs.java Example
9
public class SimpleLocalLs {public static void main(String[] args) throws Exception{
Path path = new Path("/");if ( args.length == 1){
path = new Path(args[0]);}
Configuration conf = new Configuration();FileSystem fs = FileSystem.get(conf);
FileStatus [] files = fs.listStatus(path);for (FileStatus file : files ){
System.out.println(file.getPath().getName());}
}}
Read location from the command line arguments
Acquire FileSystem Instance
List the files and directories under the provided path
Print each sub directory or file to the screen
FileSystem API: Path
• Hadoop's Path object represents a file or a directory– Not java.io.File which tightly couples to local filesystem
• Path is really a URI on the FileSystem– HDFS: hdfs://localhost/user/file1– Local: file:///user/file1
• Examples:– new Path("/test/file1.txt");– new Path("hdfs://localhost:9000/test/file1.txt");
10
Hadoop's Configuration Object
• Configuration object stores clients’ and servers’ configuration– Very heavily used in Hadoop
• HDFS, MapReduce, HBase, etc...
• Simple key-value paradigm– Wrapper for java.util.Properties class which itself is just a
wrapper for java.util.Hashtable
• Several construction options– Configuration conf1 = new Configuration();– Configuration conf2 = new Configuration(conf1);
• Configuration object conf2 is seeded with configurations of conf1 object
11
Hadoop's Configuration Object
• Getting properties is simple!• Get the property
– String nnName = conf.get("fs.default.name");• returns null if property doesn't exist
• Get the property and if doesn't exist return the provided default– String nnName = conf.get("fs.default.name",
"hdfs://localhost:9000");
• There are also typed versions of these methods:– getBoolean, getInt, getFloat, etc...– Example: int prop = conf.getInt("file.size");
12
Hadoop's Configuration Object
13
• Usually seeded via configuration files that are read from CLASSPATH (files like conf/core-site.xml and conf/hdfs-site.xml):Configuration conf = new Configuration();conf.addResource(new Path(HADOOP_HOME + "/conf/core-site.xml"));
• Must comply with the Configuration XML schema, ex:<?xml version="1.0"?><?xml-stylesheet type="text/xsl" href="configuration.xsl"?><configuration>
<property><name>fs.default.name</name><value>hdfs://localhost:9000</value>
</property></configuration>
Hadoop Configuration Object
• conf.addResource() - the parameters are either String or Path– conf.addResource("hdfs-site.xml")
• CLASSPATH is referenced when String parameter is provided
– conf.addResource(new Path("/exact/location/file.xml") :• Path points to the exact location on the local file system
• By default Hadoop loads– core-default.xml
• Located at hadoop-common-X.X.X.jar/core-default.xml
– core-site.xml
• Looks for these configuration files on CLASSPATH
14
LoadConfigurations.java Example
15
public class LoadConfigurations {private final static String PROP_NAME = "fs.default.name";
public static void main(String[] args) {Configuration conf = new Configuration();
System.out.println("After construction: " +conf.get(PROP_NAME));
conf.addResource(new Path(Vars.HADOOP_HOME +"/conf/core-site.xml"));
System.out.println("After addResource: "+conf.get(PROP_NAME));
conf.set(PROP_NAME, "hdfs://localhost:8111");System.out.println("After set: " + conf.get(PROP_NAME));
}}
1. Print the property with empty Configuration
2. Add properties from core-dsite.xml
3. Manually set the property
Run LoadConfigurations
16
$ java -cp $PLAY_AREA/HadoopSamples.jar:$HADOOP_HOME/share/hadcommon/hadoop-common-2.0.0-cdh4.0.0.jar:$HADOOP_HOME/share/hadoop/common/lib/*hdfs.LoadConfigurations
After construction: file:///After addResource: hdfs://localhost:9000After set: hdfs://localhost:8111
1. Print the property with empty Configuration
2. Add properties from core-site.xml
3. Manually set the property
FileSystem API
• Recall FileSystem is a generic abstract class used to interface with a file system
• FileSystem class also serves as a factory for concrete implementations, with the following methods– public static FileSystem get(Configuration conf)
• Will use information from Configuration such as scheme and authority
• Recall hadoop loads conf/core-site.xml by default• Core-site.xml typically sets fs.default.name property to
something like hdfs://localhost:8020– org.apache.hadoop.hdfs.DistributedFileSystem will be used by
default
– Otherwise known as HDFS17
Simple List Example
18
public class SimpleLocalLs {public static void main(String[] args) throws Exception{Path path = new Path("/");if ( args.length == 1){
path = new Path(args[0]);}Configuration conf = new Configuration();FileSystem fs = FileSystem.get(conf);FileStatus [] files = fs.listStatus(path);for (FileStatus file : files ){
System.out.println(file.getPath().getName());}
}}
What happens when you run this code?
Execute Simple List Example
• The Answer is.... it depends
19
$ java -cp $PLAY_AREA/HadoopSamples.jar:$HADOOP_HOME/share/hadoop/common/hadoop-common-2.0.0-cdh4.0.0.jar:$HADOOP_HOME/share/hadoop/common/lib/* hdfs.SimpleLslib......varsbinetc
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.SimpleLshbaselost+foundtest1tmptraininguser
●Uses java command, not yarn●core-site.xml and core-default.xml are not on the CLASSPATH●properties are then NOT added to Configuration object●Default FileSystem is loaded => local file system
●Yarn script will place core-default.xml and core-site-xml on the CLASSPATH●Properties within those files added to Configuration object●HDFS is utilized, since it was specified in core-site.xml
Reading Data from HDFS
1. Create FileSystem2. Open InputStream to a Path3. Copy bytes using IOUtils4. Close Stream
20
1: Create FileSystem
• FileSystem fs = FileSystem.get(new Configuration());– If you run with yarn command, DistributedFileSystem
(HDFS) will be created• Utilizes fs.default.name property from configuration• Recall that Hadoop framework loads core-site.xml which
sets property to hdfs (hdfs://localhost:8020)
21
2: Open Input Stream to a Path
• fs.open returns org.apache.hadoop.fs.FSDataInputStream– Another FileSystem implementation will return their own
custom implementation of InputStream
• Opens stream with a default buffer of 4k• If you want to provide your own buffer size
use– fs.open(Path f, int bufferSize)
22
...InputStream input = null;
try {input = fs.open(fileToRead);
...
3: Copy bytes using IOUtils
• Copy bytes from InputStream to OutputStream
• Hadoop’s IOUtils makes the task simple– buffer parameter specifies number of bytes to buffer at a
time
23
IOUtils.copyBytes(inputStream, outputStream, buffer);
4: Close Stream
• Utilize IOUtils to avoid boiler plate code that catches IOException
24
...} finally {
IOUtils.closeStream(input);...
ReadFile.java Example
25
public class ReadFile {public static void main(String[] args)
throws IOException {Path fileToRead = new Path("/training/data/readMe.txt");FileSystem fs = FileSystem.get(new Configuration());
InputStream input = null;try {
input = fs.open(fileToRead);
IOUtils.copyBytes(input, System.out, 4096);
} finally {IOUtils.closeStream(input);
}}
}
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.ReadFileHello from readme.txt
1: Open FileSystem
2: Open InputStream
3: Copy from Input to Output Stream
4: Close stream
Reading Data - Seek
• FileSystem.open returns FSDataInputStream– Extension of java.io.DataInputStream– Supports random access and reading via interfaces:
• PositionedReadable : read chunks of the stream• Seekable : seek to a particular position in the stream
26
Seeking to a Position
• FSDataInputStream implements Seekable interface– void seek(long pos) throws IOException
• Seek to a particular position in the file• Next read will begin at that position• If you attempt to seek past the file boundary IOException
is emitted• Somewhat expensive operation – strive for streaming and
not seeking
– long getPos() throws IOException• Returns the current position/offset from the beginning of
the stream/file
27
SeekReadFile.java Example
28
public class SeekReadFile {public static void main(String[] args) throws IOException {Path fileToRead = new Path("/training/data/readMe.txt");FileSystem fs = FileSystem.get(new Configuration());FSDataInputStream input = null;try {input = fs.open(fileToRead);System.out.print("start postion=" + input.getPos() + ": IOUtils.copyBytes(input, System.out, 4096, false);
input.seek(11);System.out.print("start postion=" + input.getPos() + ": IOUtils.copyBytes(input, System.out, 4096, false);
input.seek(0);System.out.print("start postion=" + input.getPos() + ": IOUtils.copyBytes(input, System.out, 4096, false);
} finally {IOUtils.closeStream(input);
}}
}
Start at position 0
Seek to position 11
Seek back to 0
Run SeekReadFile Example
29
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.SeekReadFile
start position=0: Hello from readme.txt
start position=11: readme.txt
start position=0: Hello from readme.txt
Write Data
1. Create FileSystem instance2. Open OutputStream
– FSDataOutputStream in this case– Open a stream directly to a Path from FileSystem– Creates all needed directories on the provided path
3. Copy data using IOUtils
30
WriteToFile.java Example
31
public class WriteToFile {public static void main(String[] args) throws IOException {String textToWrite = "Hello HDFS! Elephants are awesome!\n";InputStream in = new BufferedInputStream(
new ByteArrayInputStream(textToWrite.getBytes()));Path toHdfs = new Path("/training/playArea/writeMe.txt");Configuration conf = new Configuration();
FileSystem fs = FileSystem.get(conf);
FSDataOutputStream out = fs.create(toHdfs);
IOUtils.copyBytes(in, out, conf);}
}
1: Create FileSystem instance
2: Open OutputStream
3: Copy Data
Run WriteToFile
32
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.WriteToFile
$ hdfs dfs -cat /training/playArea/writeMe.txt
Hello HDFS! Elephants are awesome!
FileSystem: Writing Data
• Append to the end of the existing file
– Optional support by concrete FileSystem– HDFS supports
• No support for writing in the middle of the file
33
fs.append(path)
FileSystem: Writing Data
• FileSystem's create and append methods have overloaded version that take callback interface to notify client of the progress
34
FileSystem fs = FileSystem.get(conf);FSDataOutputStream out = fs.create(toHdfs, new Progressable(){
@Overridepublic void progress() {
System.out.print("..");}
});
Report progress to the screen
Overwrite Flag
• Recall FileSystem's create(Path) creates all the directories on the provided path– create(new Path(“/doesnt_exist/doesnt_exist/file/txt”)– can be dangerous, if you want to protect yourself then
utilize the following overloaded method:
35
public FSDataOutputStream create(Path f, boolean overwrite)
Set to false to make sure you do not overwrite important data
Overwrite Flag Example
36
Path toHdfs = new Path("/training/playArea/writeMe.txt");
FileSystem fs = FileSystem.get(conf); FSDataOutputStream out = fs.create(toHdfs, false);
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.BadWriteToFileException in thread "main" org.apache.hadoop.ipc.RemoteException: java.io.IOException: failed to create file /training/playArea/anotherSubDir/writeMe.txt on client 127.0.0.1 either because the filename is invalid or the file exists......
Set to false to make sure you do not overwrite important data
Error indicates that the file already exists
Copy/Move from and to Local FileSystem
• Higher level abstractions that allow you to copy and move from and to HDFS– copyFromLocalFile– moveFromLocalFile– copyToLocalFile– moveToLocalFile
37
Copy from Local to HDFS
38
FileSystem fs = FileSystem.get(new Configuration());Path fromLocal = new
Path("/home/hadoop/Training/exercises/sample_data/hamlet.txt");Path toHdfs = new Path("/training/playArea/hamlet.txt");
fs.copyFromLocalFile(fromLocal, toHdfs);
Copy file from local file system to HDFS
$ hdfs dfs -ls /training/playArea/
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.CopyToHdfs
$ hdfs dfs -ls /training/playArea/Found 1 items-rw-r—r-- hadoop supergroup /training/playArea/hamlet.txt
Empty directory before copy
File was copied
Delete Data
39
FileSystem.delete(Path path,Boolean recursive)
Path toDelete = new Path("/training/playArea/writeMe.txt");
boolean isDeleted = fs.delete(toDelete, false);System.out.println("Deleted: " + isDeleted);
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.DeleteFileDeleted: true
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.DeleteFileDeleted: false
File was already deleted by previous run
If recursive == true then non-empty directory will be deleted otherwise IOException is emitted
FileSystem: mkdirs
• Create a directory - will create all the parent directories
40
Configuration conf = new Configuration();Path newDir = new Path("/training/playArea/newDir");FileSystem fs = FileSystem.get(conf);boolean created = fs.mkdirs(newDir);System.out.println(created);
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.MkDirtrue
FileSystem: listStatus
41
Browse the FileSystem with listStatus() methods
Path path = new Path("/");Configuration conf = new Configuration();FileSystem fs = FileSystem.get(conf);FileStatus [] files = fs.listStatus(path);for (FileStatus file : files ){
System.out.println(file.getPath().getName());}
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.SimpleLstraininguser List files under ̎/̎ for HDFS
LsWithPathFilter.java example
42
FileSystem fs = FileSystem.get(conf);
FileStatus [] files = fs.listStatus(path, new PathFilter() {
@Overridepublic boolean accept(Path path) {
if (path.getName().equals("user")){return false;
}return true;
}});
for (FileStatus file : files ){System.out.println(file.getPath().getName());
}
Do not show path whose name equals to "user"
Restrict result of listStatus() by supplying PathFilter object
Run LsWithPathFilter Example
43
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.SimpleLs
traininguser
$yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.LsWithPathFilter
training
FileSystem: Globbing
• FileSystem supports file name pattern matching via globStatus() methods
• Good for traversing through a sub-set of files by using a pattern
• Support is similar to bash glob: *, ?, etc...
44
SimpleGlobbing.java
45
public class SimpleGlobbing {public static void main(String[] args)
throws IOException {Path glob = new Path(args[0]);
FileSystem fs = FileSystem.get(new Configuration());FileStatus [] files = fs.globStatus(glob);
for (FileStatus file : files ){System.out.println(file.getPath().getName());
}}
}
Read glob from command line
Similar usage to listStatus method
Run SimpleGlobbing
46
$ hdfs dfs -ls /training/data/glob/Found 4 items
drwxr-xr-x - hadoop supergroup 0 2011-12-24 11:20 /training/data/glob/2007drwxr-xr-x - hadoop supergroup 0 2011-12-24 11:20 /training/data/glob/2008drwxr-xr-x - hadoop supergroup 0 2011-12-24 11:21 /training/data/glob/2010drwxr-xr-x - hadoop supergroup 0 2011-12-24 11:21 /training/data/glob/2011
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.SimpleGlobbing /training/data/glob/201*20102011
Usage of glob with *
FileSystem: Globbing
47
Glob Explanation
? Matches any single character
* Matches zero or more characters
[abc] Matches a single character from character set {a,b,c}.
[a-b] Matches a single character from the character range {a...b}. Note that character a must be lexicographically less than or equal to character b.
[^a] Matches a single character that is not from character set or range {a}. Note that the ^ character must occur immediately to the right of the opening bracket.
\c Removes (escapes) any special meaning of character c.
{ab,cd} Matches a string from the string set {ab, cd}
{ab,c{de,fh}} Matches a string from the string set {ab, cde, cfh}
Source: FileSystem.globStatus API documentation
FileSystem
• There are several methods that return ‘true’ for success and ‘false’ for failure– delete– rename– mkdirs
• What to do if the method returns 'false'?– Check Namenode's log
• Located at $HADOOP_LOG_DIR/
48
BadRename.java
49
FileSystem fs = FileSystem.get(new Configuration());Path source = new Path("/does/not/exist/file.txt");Path nonExistentPath = new Path("/does/not/exist/file1.txt");boolean result = fs.rename(source, nonExistentPath);System.out.println("Rename: " + result);
$ yarn jar $PLAY_AREA/HadoopSamples.jar hdfs.BadRename
Rename: false
Namenode's log at $HADOOP_HOME/logs/hadoop-hadoop-namenode-hadoop-laptop.log
2011-12-25 01:18:54,684 WARN org.apache.hadoop.hdfs.StateChange: DIR* FSDirectory.unprotectedRenameTo: failed to rename /does/not/exist/file.txt to /does/not/exist/file1.txt because source does not exist
© 2012 coreservlets.com and Dima May
Customized Java EE Training: http://courses.coreservlets.com/Hadoop, Java, JSF 2, PrimeFaces, Servlets, JSP, Ajax, jQuery, Spring, Hibernate, RESTful Web Services, Android.
Developed and taught by well-known author and developer. At public venues or onsite at your location.
Wrap-Up
Summary
• We learned about– HDFS API– How to use Configuration class– How to read from HDFS– How to write to HDFS– How to browse HDFS
51
© 2012 coreservlets.com and Dima May
Customized Java EE Training: http://courses.coreservlets.com/Hadoop, Java, JSF 2, PrimeFaces, Servlets, JSP, Ajax, jQuery, Spring, Hibernate, RESTful Web Services, Android.
Developed and taught by well-known author and developer. At public venues or onsite at your location.
Questions?
JSF 2, PrimeFaces, Java 7, Ajax, jQuery, Hadoop, RESTful Web Services, Android, Spring, Hibernate, Servlets, JSP, GWT, and other Java EE training.